Recent developments in the continuous assimilation of satellite wind data for tropical cyclone track forecasting

Abstract Tropical cyclone track forecasting remains a vexing problem. Here, we have examined nine forecasts in the Australian Region which were generally deemed difficult. The forecast problem has been addressed in three ways. First, modelling and data assimilation have been performed at very high (15 km) resolution. Second, a newly developed data source, namely, high spatial and temporal resolution cloud drift winds, has been used to augment the often quite poor data base upon which forecasts are based and, in one case, these winds have been augmented with scatterometer-based surface winds. Finally, a range of continuous assimilation techniques, including recently developed 4-D variational assimilation and hourly nudging have been tested. In these cases, where often, conventional forecast guidance and CLIPER (Morison, 1993) were poor, continuous assimilation procedures, i.e. 1-hourly nudging and 4-D variational assimilation, took best advantage of the high resolution winds and provided much improved forecasts. Worthy of note was, for the case examined, scatterometer winds were able to reduce the initial position error of the cyclone. Overall, it would appear continuous assimilation, combined with a substantial high resolution data base and high resolution modelling have the potential to greatly improve the accuracy of tropical cyclone track forecasts.